- Email: [email protected]
Helix-Helix Interactions: Is the Medium the Message?
Structure
Previews Helix-Helix Interactions: Is the Medium the Message? Charles M. Deber1,2,* and Derek P. Ng1,2 1Division
of Molecular Structure and Function, Research Institute, Hospital for Sick Children, Toronto, ON M5G 1X8, Canada of Biochemistry, University of Toronto, Toronto, ON M5S 1A8, Canada *Correspondence: [email protected] http://dx.doi.org/10.1016/j.str.2015.02.004 2Department
In this issue of Structure, Zhang and colleagues compare the helix-helix interaction spaces of an extensive database of soluble and membrane proteins. Intriguingly, the resultant clusters show similar helix interaction geometries between the protein classes, differing in detail only by patterns of local interactions and interhelical distances. Protein-protein interactions are ubiquitous elements of human biology: a vast array of proteins must not only correctly fold as monomers, but then assemble or come in contact with other proteins—permanently or transiently—to elicit their function(s). Although the intraand inter-protein contacts that typify these phenomena may arise from b sheet structures found in bacterial outer membranes and mitochondria (Hutchinson et al., 1998) and/or binding to intrinsically disordered proteins (Uversky et al., 2008), a major proportion of structural and functional protein-protein interactions commonly involve a helices. In this issue of Structure, Zhang et al. (2015) analyze and compare the helixhelix interactome for soluble proteins and membrane proteins. The authors find that helical interaction surfaces in soluble proteins utilize hydrogen (H)bonding-capable residues that have larger side-chain volumes than the H-bonding residues in transmembrane (TM) helices. Additionally, the interaction surfaces in soluble proteins display a sinusoidal pattern of hydrophobicity, whereas the interaction surfaces in membrane proteins tend to contain residues with smaller side-chain volumes that serve to decrease the distances between the helices. Most generally, then, the physical/chemical factors that characterize a given helix-helix pair, whether in water or lipid—e.g., van der Waals packing, side chain-side chain, and side chain-backbone H-bonding, aromatic interactions—will accommodate the overall contributions of helix-helix, helixsolvent, and solvent-solvent interactions (White and Wimley, 1999). Principal as-
pects of helix-helix interactions are shown schematically in Figure 1. Perhaps the most important advance in the Zhang et al. (2015) paper is the codifying of the membrane protein helix-helix interactome and, in the process, raising our appreciation of the a helix as the ideal structure for spanning a cellular membrane. The helix is linear, thereby minimizing disruption of lipid packing; the electrostatic aspects of the peptide bonds are compensated by internal H bonds, and the helix side chains project directly into the lipid, where the array of hydrophobic side chains will ostensibly be compatible with the bilayer core. There is, however, a major additional consideration that may not be immediately apparent from the linear sequence, viz., folding of a linear protein chain into a helix creates ‘‘faces’’—in effect an asymmetry of residues that is defined by the repeat distance turn of the helix; thus, given the periodicity of 3.6 residues/turn, residues 1 and 5 are on the same face, whereas residues 3 and 7 will occupy an opposing face. It is this latter effect that not only governs helix-helix interactions per se, but also provides the challenge to researchers in the field: given two helices of known sequence and sufficient length (18–20 residues; about four to five turns of helix) to span a bilayer, what will be the preferred interface when they interact? These considerations bear on the observation that in TM helices, ‘‘small’’ (Gly/Ala/Ser) and/or ‘‘large’’ (Ile/ Val/Leu) residues are engaged in ‘‘knobs-into-holes’’ packing in helix-helix contacts that maximize van der Waals interactions, as exemplified by intercalation of the Gly ‘‘holes’’ and Val ‘‘knobs’’ in the
GG4 motif of the human erythrocyte glycoprotein glycophorin A (GpA) dimer (MacKenzie et al., 1997). How do the forces that stabilize these helix-helix interactions differ between soluble and membrane protein helices? Broadly stated, in the absence of the ‘‘hydrophobic effect’’—the minimization of hydrophobic residue contact with the aqueous medium—globular proteins would not be driven to fold into their functional secondary and tertiary structures. This effect ensures that the hydrophobic components of soluble proteins are largely relegated to the protein ‘‘core’’ to satisfy water’s relative inability to solvate their non-polar side chains, with water instead solvating the defaulted polar protein surface. The hydrophobic effect is further driven by entropic considerations, as water ‘‘structured’’ around hydrophobic protein components is concomitantly minimized. Interestingly, membrane protein biosynthesis is similarly characterized by a manifestation of the hydrophobic effect: nascent hydrophobic protein sequences are partitioned from the aqueous channel of the translocon into the receptive environment of the lipid bilayer (Hessa et al., 2005). But what is hydrophobicity in a membrane? Other than propelling the low percentage of polar residues (< 20%; often Ser, Thr, and Gln) into helix-helix interfaces (which one may term a ‘‘lipophobic effect’’), a helix with largely hydrophobic residues around its circumference should, in principle, be otherwise compatible with the membranous milieu. Yet, if this were the final situation, helices would not be driven to pack and fold into the dimeric and multimeric
Structure 23, March 3, 2015 ª2015 Elsevier Ltd All rights reserved 437
Structure
Previews
(Bellmann-Sickert et al., 2015). Given that a single point mutation at a helix-helix interface can propagate deleterious effects on folding and assembly (Ng et al., 2012), the present categorization of the principles and motifs that underlie helixhelix interactions will be of great value in disease- and mutant-related scenarios in which intervention in protein-protein interactions would be desirable. ACKNOWLEDGMENTS Support was provided by the Canadian Institutes of Health Research (CIHR FRN-5810). REFERENCES
Figure 1. Helix-Helix Interactions in Proteins Folding via helix-helix pairing depends on a balance of factors: helix-helix, helix-solvent, and solventsolvent interactions, as highlighted by the circles in the diagram. For soluble proteins, the solvent is the water surrounding the folded protein (shown in blue, with some water molecules indicated). For native membrane proteins, the solvent is the bilayer acyl chains (shown in white). Local interactions of the non-interfacial residues of the helix-helix pair with other protein substituents (not shown) are further structural determinants of the system. Gray ribbons represent the protein backbone. Yellow and purple spheres represent amino acid side chains from each of the two helices.
structures common to membrane proteins, but would rather remain completely surrounded by lipids. Thus, not only will the irregular topology of a TM helix impact the acyl chain packing of the bilayer per se, but one must also consider the subtleties of the resulting lipid-helix interactions. For example, it has been shown that helixhelix interaction sites in membrane proteins are those most poorly solvated by lipid acyl chains (Johnson et al., 2006). Thus, similar to the ‘‘poor solvent’’ role of water in soluble protein folding, the inability of the extended lipid acyl CH2 units to intercalate effectively with specific regions of the TM sequence may help define TM helix packing and assembly. In this context, Zhang et al. (2015) find that pairs of membrane-embedded helices are closer together than their soluble
counterparts, indicating that, in the absence of the hydrophobic effect, a combination of forces unique to the membrane environment acts to hold TM helices together. Knowledge of the factors that contribute to helix-helix interactions facilitates the pathway to the design of protein-protein interaction ‘‘disruptors’’— reagents that target protein assembly motifs (Higueruelo et al., 2013). For example, synthetic peptides with native TM sequences can disrupt ErbB2 receptor dimerization and abolish aberrant signaling that may play a role in cancer (Bennasroune et al., 2004). Similarly, peptides complementary to a GG7 helical heptad motif inhibit drug efflux by disrupting the helix-helix interaction site of bacterial multidrug resistance proteins
438 Structure 23, March 3, 2015 ª2015 Elsevier Ltd All rights reserved
Bellmann-Sickert, K., Stone, T.A., Poulsen, B.E., and Deber, C.M. (2015). J. Biol. Chem. 290, 1752–1759. Bennasroune, A., Fickova, M., Gardin, A., DirrigGrosch, S., Aunis, D., Cre´mel, G., and Hubert, P. (2004). Mol. Biol. Cell 15, 3464–3474. Hessa, T., Kim, H., Bihlmaier, K., Lundin, C., Boekel, J., Andersson, H., Nilsson, I., White, S.H., and von Heijne, G. (2005). Nature 433, 377–381. Higueruelo, A.P., Jubb, H., and Blundell, T.L. (2013). Curr. Opin. Pharmacol. 13, 791–796. Hutchinson, E.G., Sessions, R.B., Thornton, J.M., and Woolfson, D.N. (1998). Protein Sci. 7, 2287– 2300. Johnson, R.M., Rath, A., Melnyk, R.A., and Deber, C.M. (2006). Biochemistry 45, 8507–8515. MacKenzie, K.R., Prestegard, J.H., and Engelman, D.M. (1997). Science 276, 131–133. Ng, D.P., Poulsen, B.E., and Deber, C.M. (2012). Biochim. Biophys. Acta 1818, 1115–1122. Uversky, V.N., Oldfield, C.J., and Dunker, A.K. (2008). Annu. Rev. Biophys. 37, 215–246. White, S.H., and Wimley, W.C. (1999). Annu. Rev. Biophys. Biomol. Struct. 28, 319–365. Zhang, S.-Q., Kulp, D.W., Schramm, C.A., Mravic, M., Samish, I., and DeGrado, W.F. (2015). Structure 23, this issue, 527–541.
Previews Helix-Helix Interactions: Is the Medium the Message? Charles M. Deber1,2,* and Derek P. Ng1,2 1Division
of Molecular Structure and Function, Research Institute, Hospital for Sick Children, Toronto, ON M5G 1X8, Canada of Biochemistry, University of Toronto, Toronto, ON M5S 1A8, Canada *Correspondence: [email protected] http://dx.doi.org/10.1016/j.str.2015.02.004 2Department
In this issue of Structure, Zhang and colleagues compare the helix-helix interaction spaces of an extensive database of soluble and membrane proteins. Intriguingly, the resultant clusters show similar helix interaction geometries between the protein classes, differing in detail only by patterns of local interactions and interhelical distances. Protein-protein interactions are ubiquitous elements of human biology: a vast array of proteins must not only correctly fold as monomers, but then assemble or come in contact with other proteins—permanently or transiently—to elicit their function(s). Although the intraand inter-protein contacts that typify these phenomena may arise from b sheet structures found in bacterial outer membranes and mitochondria (Hutchinson et al., 1998) and/or binding to intrinsically disordered proteins (Uversky et al., 2008), a major proportion of structural and functional protein-protein interactions commonly involve a helices. In this issue of Structure, Zhang et al. (2015) analyze and compare the helixhelix interactome for soluble proteins and membrane proteins. The authors find that helical interaction surfaces in soluble proteins utilize hydrogen (H)bonding-capable residues that have larger side-chain volumes than the H-bonding residues in transmembrane (TM) helices. Additionally, the interaction surfaces in soluble proteins display a sinusoidal pattern of hydrophobicity, whereas the interaction surfaces in membrane proteins tend to contain residues with smaller side-chain volumes that serve to decrease the distances between the helices. Most generally, then, the physical/chemical factors that characterize a given helix-helix pair, whether in water or lipid—e.g., van der Waals packing, side chain-side chain, and side chain-backbone H-bonding, aromatic interactions—will accommodate the overall contributions of helix-helix, helixsolvent, and solvent-solvent interactions (White and Wimley, 1999). Principal as-
pects of helix-helix interactions are shown schematically in Figure 1. Perhaps the most important advance in the Zhang et al. (2015) paper is the codifying of the membrane protein helix-helix interactome and, in the process, raising our appreciation of the a helix as the ideal structure for spanning a cellular membrane. The helix is linear, thereby minimizing disruption of lipid packing; the electrostatic aspects of the peptide bonds are compensated by internal H bonds, and the helix side chains project directly into the lipid, where the array of hydrophobic side chains will ostensibly be compatible with the bilayer core. There is, however, a major additional consideration that may not be immediately apparent from the linear sequence, viz., folding of a linear protein chain into a helix creates ‘‘faces’’—in effect an asymmetry of residues that is defined by the repeat distance turn of the helix; thus, given the periodicity of 3.6 residues/turn, residues 1 and 5 are on the same face, whereas residues 3 and 7 will occupy an opposing face. It is this latter effect that not only governs helix-helix interactions per se, but also provides the challenge to researchers in the field: given two helices of known sequence and sufficient length (18–20 residues; about four to five turns of helix) to span a bilayer, what will be the preferred interface when they interact? These considerations bear on the observation that in TM helices, ‘‘small’’ (Gly/Ala/Ser) and/or ‘‘large’’ (Ile/ Val/Leu) residues are engaged in ‘‘knobs-into-holes’’ packing in helix-helix contacts that maximize van der Waals interactions, as exemplified by intercalation of the Gly ‘‘holes’’ and Val ‘‘knobs’’ in the
GG4 motif of the human erythrocyte glycoprotein glycophorin A (GpA) dimer (MacKenzie et al., 1997). How do the forces that stabilize these helix-helix interactions differ between soluble and membrane protein helices? Broadly stated, in the absence of the ‘‘hydrophobic effect’’—the minimization of hydrophobic residue contact with the aqueous medium—globular proteins would not be driven to fold into their functional secondary and tertiary structures. This effect ensures that the hydrophobic components of soluble proteins are largely relegated to the protein ‘‘core’’ to satisfy water’s relative inability to solvate their non-polar side chains, with water instead solvating the defaulted polar protein surface. The hydrophobic effect is further driven by entropic considerations, as water ‘‘structured’’ around hydrophobic protein components is concomitantly minimized. Interestingly, membrane protein biosynthesis is similarly characterized by a manifestation of the hydrophobic effect: nascent hydrophobic protein sequences are partitioned from the aqueous channel of the translocon into the receptive environment of the lipid bilayer (Hessa et al., 2005). But what is hydrophobicity in a membrane? Other than propelling the low percentage of polar residues (< 20%; often Ser, Thr, and Gln) into helix-helix interfaces (which one may term a ‘‘lipophobic effect’’), a helix with largely hydrophobic residues around its circumference should, in principle, be otherwise compatible with the membranous milieu. Yet, if this were the final situation, helices would not be driven to pack and fold into the dimeric and multimeric
Structure 23, March 3, 2015 ª2015 Elsevier Ltd All rights reserved 437
Structure
Previews
(Bellmann-Sickert et al., 2015). Given that a single point mutation at a helix-helix interface can propagate deleterious effects on folding and assembly (Ng et al., 2012), the present categorization of the principles and motifs that underlie helixhelix interactions will be of great value in disease- and mutant-related scenarios in which intervention in protein-protein interactions would be desirable. ACKNOWLEDGMENTS Support was provided by the Canadian Institutes of Health Research (CIHR FRN-5810). REFERENCES
Figure 1. Helix-Helix Interactions in Proteins Folding via helix-helix pairing depends on a balance of factors: helix-helix, helix-solvent, and solventsolvent interactions, as highlighted by the circles in the diagram. For soluble proteins, the solvent is the water surrounding the folded protein (shown in blue, with some water molecules indicated). For native membrane proteins, the solvent is the bilayer acyl chains (shown in white). Local interactions of the non-interfacial residues of the helix-helix pair with other protein substituents (not shown) are further structural determinants of the system. Gray ribbons represent the protein backbone. Yellow and purple spheres represent amino acid side chains from each of the two helices.
structures common to membrane proteins, but would rather remain completely surrounded by lipids. Thus, not only will the irregular topology of a TM helix impact the acyl chain packing of the bilayer per se, but one must also consider the subtleties of the resulting lipid-helix interactions. For example, it has been shown that helixhelix interaction sites in membrane proteins are those most poorly solvated by lipid acyl chains (Johnson et al., 2006). Thus, similar to the ‘‘poor solvent’’ role of water in soluble protein folding, the inability of the extended lipid acyl CH2 units to intercalate effectively with specific regions of the TM sequence may help define TM helix packing and assembly. In this context, Zhang et al. (2015) find that pairs of membrane-embedded helices are closer together than their soluble
counterparts, indicating that, in the absence of the hydrophobic effect, a combination of forces unique to the membrane environment acts to hold TM helices together. Knowledge of the factors that contribute to helix-helix interactions facilitates the pathway to the design of protein-protein interaction ‘‘disruptors’’— reagents that target protein assembly motifs (Higueruelo et al., 2013). For example, synthetic peptides with native TM sequences can disrupt ErbB2 receptor dimerization and abolish aberrant signaling that may play a role in cancer (Bennasroune et al., 2004). Similarly, peptides complementary to a GG7 helical heptad motif inhibit drug efflux by disrupting the helix-helix interaction site of bacterial multidrug resistance proteins
438 Structure 23, March 3, 2015 ª2015 Elsevier Ltd All rights reserved
Bellmann-Sickert, K., Stone, T.A., Poulsen, B.E., and Deber, C.M. (2015). J. Biol. Chem. 290, 1752–1759. Bennasroune, A., Fickova, M., Gardin, A., DirrigGrosch, S., Aunis, D., Cre´mel, G., and Hubert, P. (2004). Mol. Biol. Cell 15, 3464–3474. Hessa, T., Kim, H., Bihlmaier, K., Lundin, C., Boekel, J., Andersson, H., Nilsson, I., White, S.H., and von Heijne, G. (2005). Nature 433, 377–381. Higueruelo, A.P., Jubb, H., and Blundell, T.L. (2013). Curr. Opin. Pharmacol. 13, 791–796. Hutchinson, E.G., Sessions, R.B., Thornton, J.M., and Woolfson, D.N. (1998). Protein Sci. 7, 2287– 2300. Johnson, R.M., Rath, A., Melnyk, R.A., and Deber, C.M. (2006). Biochemistry 45, 8507–8515. MacKenzie, K.R., Prestegard, J.H., and Engelman, D.M. (1997). Science 276, 131–133. Ng, D.P., Poulsen, B.E., and Deber, C.M. (2012). Biochim. Biophys. Acta 1818, 1115–1122. Uversky, V.N., Oldfield, C.J., and Dunker, A.K. (2008). Annu. Rev. Biophys. 37, 215–246. White, S.H., and Wimley, W.C. (1999). Annu. Rev. Biophys. Biomol. Struct. 28, 319–365. Zhang, S.-Q., Kulp, D.W., Schramm, C.A., Mravic, M., Samish, I., and DeGrado, W.F. (2015). Structure 23, this issue, 527–541.